Catch up on select AI news and developments from the past week or so:
Google ships 'Personal Intelligence' as a deeply personalized assistant across its ecosystem. Google has introduced Personal Intelligence for AI Mode in Search and Gemini, allowing the assistant (with user permission) to connect across Gmail, Photos, Search history, YouTube, and more to answer questions with highly specific personal context. The system can infer preferences and history from account "breadcrumbs," retrieve details like insurance renewal timing from email, and tailor travel plans based on family context. Google says it filters or obfuscates personal data in conversations while still using account signals to locate requested information.
Importance for marketers: This raises the bar for personalization and makes data governance a brand issue. Expect higher user sensitivity to consent, transparency, and value exchange, plus new competitive pressure to deliver genuinely helpful experiences without crossing privacy lines.
Retailers examine options for on-AI retail. Major retailers, including Walmart, Target and Etsy, are expanding into AI-led commerce by integrating with platforms such as ChatGPT, Gemini, and Copilot, allowing purchases directly within AI conversations. AI-driven traffic to US ecommerce sites surged 758% year over year in November 2025, with Cyber Monday seeing a 670% increase in AI-referred visits. While agentic commerce expands reach, retailers risk surrendering data ownership and brand control as discovery, evaluation, and checkout increasingly occur inside third-party AI platforms rather than owned channels.
Importance for marketers: AI is rapidly becoming a primary commerce interface. Marketers must rethink channel strategy, data ownership, attribution, and brand differentiation as AI agents mediate customer relationships and potentially reduce loyalty to direct brand experiences.
Google's AI Overviews upgrade: Gemini 3 powers smoother chat handoffs. Google has made Gemini 3 the default model powering AI Overviews globally and introduced seamless transitions into AI Mode conversations. The update allows users to move from summary results into contextual follow-up chats without losing continuity. AI Mode now serves 75 million daily active users across 40 languages. The shift positions Search as a fluid, conversational environment rather than a click-based interface, further embedding AI directly into user research journeys and long-form information exploration.
Importance for marketers: Search behavior is becoming conversational and persistent. Marketers must optimize for AI-mediated discovery, contextual follow-ups and entity-based visibility rather than traditional click-through SEO alone.
Yahoo launches Scout, an AI answer engine built on user data and Web grounding. Yahoo has debuted Scout, an AI-powered answer engine in US beta designed to convert queries into direct answers across tasks like shopping comparisons, stock analysis, weather planning, and fact checking. The product claims to use hundreds of millions of user profiles and a large entity knowledge graph, with Claude as its primary model and Bing's grounding API to support citations from the open web. Scout is positioned as an alternative to conversational search products and is expected to expand features and personalization over time.
Importance for marketers: Another scaled answer engine means more competition for attention above the click. Marketers should prepare for multi-engine optimization, stronger emphasis on entity credibility, and new shopping surfaces where "best answer" dynamics can override brand preference.
UK proposes forcing Google to let publishers opt out of AI summaries. The UK Competition and Markets Authority has proposed requiring Google to allow publishers to opt out of AI Overviews content scraping, increase transparency and properly cite sources. Publishers report traffic declines following the rollout of AI summaries that reduce click-through behavior. The proposal also recommends fair search ranking practices and default search choice screens. The consultation closes February 25 and could materially alter how AI-generated search summaries operate within the UK market.
Importance for marketers: Regulatory pressure may reshape AI search economics. Brands relying on organic visibility must monitor opt-out policies, attribution standards and regional compliance shifts that could influence traffic and content strategy.
Facing AI slop and a trust problem, AI platforms invest in Super Bowl-level brand ads. Major AI platforms including OpenAI and Meta are investing heavily in brand advertising, including Super Bowl placements, to combat rising consumer distrust and backlash against AI-generated content. Research shows a perception gap: while 82% of advertising executives believe younger consumers feel positively about AI ads, only 45% of those consumers agree. Transparency frameworks are emerging as trust declines amid concerns around deepfakes, hallucinations and AI-driven personalization.
Importance for marketers: AI brands are now behaving like consumer brands. Trust, narrative control, and transparency labeling are becoming strategic imperatives in both enterprise and consumer AI marketing.
Scammers exploit OpenAI's name in fake ad platform targeting marketers. A phishing campaign is targeting marketers and app testers with fake TestFlight invitations for a non-existent product called OpenAI Advertising GPT. The scam promises advertising credits and AI-powered campaign tools, harvesting email and usage data while exploiting OpenAI's legitimate plans to test advertising within ChatGPT. The timing aligns closely with real product announcements, increasing credibility and risk for marketing professionals seeking early access opportunities.
Importance for marketers: As AI ad platforms emerge, brand impersonation risk rises. Marketing teams must implement stricter verification protocols for beta invitations and vendor outreach to prevent data compromise.
Meta outlines a shift toward AI-generated, personalized social feeds and monetization. Meta leadership is describing AI as the next major media format, moving feeds toward experiences that are more immersive, interactive, and increasingly generated or remixed by AI. The company has already introduced a feed of short AI-generated videos inside its AI app and is teasing prompt-built worlds or games that users can share. The direction aligns with plans to monetize its chatbot through subscriptions and advertising, while VR investments continue to operate alongside the AI roadmap.
Importance for marketers: Content supply will explode while authenticity signals get murkier. Brand safety, creative differentiation, and measurement will become harder as AI-generated media blends with human content and platforms introduce new ad and subscription layers around assistants.
Claude adds interactive workplace apps for Slack, Canva, Figma, Box, and more. Claude now supports interactive apps inside the chat experience, starting with workplace tools such as Slack, Canva, Figma, Box, and Clay, with a Salesforce implementation expected. The setup allows a logged-in version of the service to be accessible from within Claude so users can perform tasks like sending messages, generating charts, or accessing files, depending on enabled apps. The feature is available to paid tiers and is built on MCP, enabling an ecosystem of third-party tool integrations.
Importance for marketers: This pulls marketing work into an agent-plus-tool workflow. Expect faster creative iteration, more "chat-to-execution" experiences, and new governance needs around permissions, auditability, and brand asset control across design and collaboration platforms.
ServiceNow makes Claude the default model for building agentic workflows. ServiceNow has set Claude as the default model for its Build Agent and is extending it across additional development workflows while still supporting model choice. The plan emphasizes governed orchestration, observability, and lifecycle controls through an AI Control Tower, alongside industry-focused agentic workflows for domains like healthcare and life sciences. ServiceNow is also collaborating with another model provider for automation and speech-to-speech capabilities, signaling a multi-model approach where different systems handle different tasks.
Importance for marketers: Enterprise buyers want agents that are governed, measurable, and safe. Marketing ops and rev ops teams should expect stronger demand for audit trails, permissioning, and predictable costs before assistants are trusted with customer-facing actions.
Congress proposes subpoena rights for AI-trained content. The Transparency and Responsibility for Artificial Intelligence Networks Act would establish an administrative subpoena process allowing copyright owners to determine whether generative AI models were trained on their protected works. The bipartisan bill requires AI developers to disclose training materials upon request, creating a new legal mechanism for transparency. The proposal arrives amid escalating lawsuits and settlements involving AI training data and could materially affect how models are built and documented.
Importance for marketers: Content sourcing, licensing and AI training transparency are becoming compliance issues. Brands using AI-generated content must monitor legal developments that may influence data provenance and intellectual property risk.
Franny Hsiao, Salesforce: Scaling enterprise AI. Enterprise AI initiatives often fail due to weak data architecture rather than model performance. Salesforce emphasizes production-grade governance, observability and human-in-the-loop safeguards as essential for scaling. Techniques such as streaming responses to reduce perceived latency, edge-based offline AI for field operations and standardized agent protocols aim to improve reliability and interoperability. The future bottleneck, according to Salesforce, lies in preparing enterprise data to be agent-ready rather than in model capability alone.
Importance for marketers: Successful AI adoption depends on infrastructure, governance, and usability. Marketing leaders deploying AI agents must invest in observability, data quality, and workflow design rather than focusing solely on model selection.
Google rolls out a lower-cost AI Plus plan globally, including the US. Google has expanded its more affordable AI Plus subscription to all markets where its AI plans are available, pricing it at $7.99 per month in the US. The plan includes access to Gemini 3 Pro, Nano Banana Pro, creative tools, and research/writing support, plus 200GB of storage and family sharing. Some existing subscribers will gain access automatically. The move broadens the addressable market for paid AI features by offering an entry tier below the higher-priced plan.
Importance for marketers: Lower pricing can accelerate consumer and SMB AI adoption, increasing AI-mediated discovery and creation. Marketers should watch which features are bundled at each tier to anticipate changes in content production volume and user expectations.
Lenovo plans a multi-model approach to AI across PCs, mobiles, and wearables. Lenovo is aiming to integrate AI broadly across its product portfolio and is pursuing partnerships with multiple large language models across regions rather than building its own. Its cross-device intelligence layer is designed to integrate with different model partners to account for varied regulatory environments and market needs. The company expects rising component costs and plans to pass increases to customers, while also pursuing infrastructure partnerships that can speed deployment of data center capability globally.
Importance for marketers: Multi-model device ecosystems can fragment default assistant behavior by geography and hardware. Marketers should anticipate inconsistent user experiences and plan for messaging, support, and measurement that works across multiple assistants and device contexts.
WhatsApp will now charge AI chatbots to operate in Italy. Meta will begin charging developers to operate AI chatbots on WhatsApp in Italy, where regulators required the company to allow third-party bots. Developers will pay per AI-generated message, potentially generating substantial costs at scale. The move follows broader regulatory scrutiny across the EU and Brazil and could establish pricing precedents for other regions. Developers must now redirect users off-platform or absorb new API costs.
Importance for marketers: Platform economics for conversational AI are tightening. Brands deploying chatbots via messaging platforms must model rising API costs and regulatory-driven policy shifts in customer engagement strategy.
Gallup Workforce shows details of AI adoption in US workplaces. Gallup survey data from over 23,000 US workers shows AI adoption remains concentrated among knowledge workers in tech, finance, and professional services. Only about 10% of employees use AI daily, while nearly a quarter are unsure whether their employer has adopted AI at all. Usage centers on information consolidation, search, and idea generation, with limited embedding into daily workflows. Adoption gaps are wider among non-managerial and frontline staff.
Importance for marketers: AI literacy and adoption remain uneven inside organizations. Marketing leaders should prioritize internal enablement and communication to ensure tools are understood and consistently integrated into workflows.
Runway research shows most people cannot reliably spot AI-generated video. Runway published research and a public test experience suggesting that participants struggled to distinguish real five-second videos from AI-generated outputs created by its latest model. In a study of over 1,000 participants judging a mix of real and generated clips, only a small minority achieved statistically significant accuracy, with overall performance only slightly above chance. Results varied by content type, and the work argues that detection alone is not a sufficient trust mechanism, pointing to provenance standards like C2PA.
Importance for marketers: Synthetic video is crossing the believability threshold. Expect higher stakes for disclosure, brand safety, and crisis response, plus growing pressure to adopt provenance standards and clear labeling as audiences lose confidence in visual proof.
Open-source personal assistant Clawdbot goes viral, then rebrands to Moltbot. A viral open-source personal AI assistant that runs locally has attracted attention for offering agentic behavior, persistent memory, and deep access to user tools like email, calendars, and messaging apps. The tool has been renamed Moltbot after trademark pressure tied to similarity with the name of major model family Claude. The project highlights both the appeal of locally run agents and the risks of granting an AI shell access and broad tool permissions, with security guidance emphasizing threat models and cautious setup.
Importance for marketers: Local agents expand the market for assistants that can act, not just chat. They also raise security and brand trust concerns: marketers should harden internal policies on tool access, permissions, and vendor evaluation as DIY agents enter workplaces.
Models that improve on their own are AI's next big thing. Researchers and leading AI labs are exploring recursive self-improvement, enabling models to refine themselves after deployment. Advocates argue this could accelerate AI research and development, while critics warn it introduces harder-to-detect risks and governance challenges. As models increasingly participate in AI R&D pipelines, policymakers face limited visibility into automation processes. The technique recalls earlier successes in game-based systems but faces far greater complexity in real-world applications.
Importance for marketers: Self-improving models could accelerate product cycles and capabilities. However, governance, transparency and risk oversight will become increasingly important as AI systems gain greater autonomy.
DeepSeek updates OCR using an Alibaba model and open-sources the system. DeepSeek has introduced an updated OCR system that swaps its prior approach for an Alibaba Qwen-based model and a new encoder design intended to improve document understanding and reading order. The write-up claims measurable benchmark gains and improved efficiency by compressing complex pages into fewer visual tokens while preserving accuracy. The release is open-sourced, enabling developers to adopt or extend the technology for document-heavy workflows such as contracts, forms, and records where layout, structure, and extraction accuracy matter.
Importance for marketers: Better OCR improves content reuse and analytics for scanned assets like PDFs, contracts, and legacy collateral. It can speed enablement, compliance workflows, and campaign operations, especially when teams need structured data from messy documents.
Bill Gates and tech leaders reflect on AI fears. Prominent tech leaders are expressing concern about AI's long-term impact on meaning, employment and human agency. Reflections center on the psychological implications of automation and the possibility that AI could displace not only tasks but purpose-driven work. While some emphasize human choice and adaptability, others warn about societal disruption if superhuman intelligence advances without adequate safeguards.
Importance for marketers: Cultural sentiment toward AI is evolving beyond functionality into existential concerns. Brand messaging and product positioning must consider human meaning, agency and trust in AI-driven futures.
'Jobs, jobs, jobs' is the AI mantra in Davos, but skeptics are unconvinced. Executives at the World Economic Forum in Davos emphasized that artificial intelligence will ultimately create more jobs than it eliminates, repeating a familiar refrain of "jobs, jobs, jobs" to counter mounting public anxiety. Business leaders highlighted productivity gains, new roles, and long-term economic expansion driven by AI adoption. Yet skepticism persists, particularly among labor advocates and policy observers who question the pace of workforce retraining and the near-term disruption to white-collar and knowledge work. While optimism dominated executive messaging, concerns around inequality and displacement remain unresolved.
Importance for marketers: Workforce anxiety shapes brand perception. Marketing leaders must consider how AI positioning affects employer branding, customer trust, and public sentiment in a climate where economic disruption fears are intensifying.
Google adds Nano Banana and other AI features to Chrome. Google is embedding Gemini 3-powered features into Chrome, including the Nano Banana image editor, an AI side-panel assistant and auto-browse capabilities for subscribers. Users can shop, log into websites and draft social posts with AI assistance, though final approval remains manual. The move further integrates AI directly into browsing behavior, reinforcing Chrome's role as an AI-native interface amid competition from AI-focused browsers.
Importance for marketers: Browsers are evolving into AI agents. Discovery, content creation and ecommerce interactions may increasingly occur within AI-enhanced browser environments, reshaping user journey assumptions.
Trump's use of AI images pushes new boundaries, further eroding public trust. The use of AI-generated and altered political imagery by official government channels has raised concerns among misinformation experts about eroding public trust. The blending of memes and realistic synthetic media complicates distinctions between satire and manipulation. Experts warn that widespread sharing of unlabeled AI-edited content by credible institutions may accelerate distrust and normalize synthetic media in high-stakes contexts.
Importance for marketers: Synthetic media normalization increases reputational risk. Brands must consider labeling standards, authenticity safeguards and media literacy in an environment where realistic AI-generated content blurs truth boundaries.
Microsoft adds AI features to Paint and Notepad, including coloring-book generation. Microsoft is rolling out AI upgrades to lightweight Windows apps in insider channels. Paint gains a "Coloring book" feature that generates line-art templates from text prompts, outputting multiple options users can place on a canvas for printing or digital coloring, with some features limited to Copilot Plus PCs. Notepad adds streaming previews for write, rewrite, and summarize capabilities, plus interface onboarding and additional Markdown support. The additions suggest experimentation in making everyday utilities part of the AI value story.
Importance for marketers: AI is becoming ambient in default software, broadening casual creation. Expect more consumer-generated creative assets, more "promptable" micro-content, and increased need for brand guidance on what is acceptable, accurate, and rights-safe.
xAI's Grok Imagine now generates 10-second AI videos. xAI has expanded Grok Imagine to generate 10-second AI videos, extending its multimodal capabilities beyond images into short-form motion content. The update allows users to create brief animated clips from prompts, intensifying competition in the generative video market alongside offerings from major AI labs. As text-to-video tools become more accessible and integrated into consumer-facing platforms, realistic synthetic content is becoming easier to produce at scale. The development underscores how rapidly video generation quality and usability are advancing across the AI landscape.
Importance for marketers: Short-form generative video lowers creative barriers and accelerates content production. Brands must prepare for increased synthetic media volume, stronger disclosure expectations, and new opportunities to prototype campaign visuals quickly.
You can find the previous issue of AI Update here.
Editor's note: GPT-5.2 was used to help compile this issue of AI Update.
